Background on the Opioid Crisis in America

The United States opioid epidemic is a nationwide public health crisis. Initially driven by increased consumption and availability of pharmaceutical opioids such as oxycontin, an increasing number of opioid overdoses are now related to heroin and illicitly manufactured fentanyl and fentanyl analogs. Deaths related to opioid use have sharply increased from 1996 when advertisement around opioid based pharmaceuticals became unrestricted.

Research Questions

I wanted to look at the steady increase in opioid-related deaths in the US, to visualize this I used a simple line graph plotting the data from 1990 to 2019. I also compared opioid related deaths to other drug related deaths in the US along the same time scale, initially just alcohol but then also cocaine and amphetamine. As my final research question I wanted to plot the data spatially by looking at how opioid related deaths have differed across different US states. However, I wanted to keep the temporal aspect of the data, so therefore I made a data file for every year with every states opioid-related death rate

Data Sources

All the data was sourced from the Institute for Health Metrics and Evaluation

Alcohol vs. Opioid- related Deaths

Data preparation

#Creating line graph for opioid  and alcohol related deaths over time 1990-2019 in the USA

#install and load relevant packages

library(dplyr)
library(here)
library(tidyverse)
library(readr)
library(ggplot2)

#load data
usdeathcompared <- read.csv(here('alcoholvsopioid.csv'))

#deleting unnecessary columns

columns_to_delete <- c("measure_id", "measure_name","cause_id", "metric_id","metric_name","age_id","sex_id","sex_name","age_name","upper","lower", "location_name")
usdeathcompared <- usdeathcompared[, !(names(usdeathcompared) %in% columns_to_delete)]

#renaming columns

##val change to rate_deaths 
colnames(usdeathcompared)[colnames(usdeathcompared)=="val"] <-"rate_death"
view(usdeathcompared)

#make line graph 

#make wide data

usdeath<- usdeathcompared %>%
  pivot_wider(names_from = cause_name, values_from = rate_death)

#change column names
colnames(usdeath)[colnames(usdeath)=="Alcohol use disorders"] <-"Alcohol"
colnames(usdeath)[colnames(usdeath)=="Opioid use disorders"] <-"Opioid"

#plot data

alcoholvsopioid <- ggplot(usdeath, aes(x = year)) +
  geom_line(aes(y = Alcohol, color = factor("Alcohol")), linetype = "solid", size = 1.65) +
  geom_line(aes(y = Opioid, color = factor("Opioid")), linetype = "solid", size = 1.65) +
  scale_color_manual(name = "Cause", values = c("Alcohol" = "dodgerblue", "Opioid" = "firebrick")) +
  labs(title = "Alcohol- and Opioid Use-Related Death Rates", 
       subtitle =  "US Data from 1990-2019",
       caption = "Source: Institute for Health Metrics and Evaluation",
       x = "Year",
       y = "Death Rate per 100,000",
       color = "Cause") +
  scale_y_continuous(limits = c(0, max(usdeath$Alcohol, usdeath$Opioid) + 1)) +
  theme_minimal() +
  theme(plot.title = element_text(color = "#0099f9", size = 20, face = "bold", hjust = 0.5),
  plot.subtitle = element_text(size = 13, face = "bold", hjust = 0.5),
  plot.caption = element_text(face = "italic", hjust = 0)) +
  theme(axis.title.x = element_text(color = "#0099f9", size = 16, face = "bold"),
        axis.title.y = element_text(color = "#0099f9", size = 16, face = "italic"))

##Interactive Plot Next I decided I wanted to make an interactive plot for the main 4 drug related deaths in the US

library(here)
library(tidyverse)
library(readr)
library(gganimate)
## Warning: package 'gganimate' was built under R version 4.3.3
library(plotly)
## Warning: package 'plotly' was built under R version 4.3.3
## 
## Attaching package: 'plotly'
## The following object is masked from 'package:ggplot2':
## 
##     last_plot
## The following object is masked from 'package:stats':
## 
##     filter
## The following object is masked from 'package:graphics':
## 
##     layout
usdrugdeaths <- read.csv(here('alldrugsus.csv'))


#deleting unnessecary columns

columns_to_delete <- c("measure_id", "measure_name","cause_id", "metric_id","metric_name","age_id","sex_id","sex_name","age_name","upper","lower", "location_name")
usdrugdeaths <- usdrugdeaths[, !(names(usdrugdeaths) %in% columns_to_delete)]

#renaming columns

##val change to rate_deaths 
colnames(usdrugdeaths)[colnames(usdrugdeaths)=="val"] <-"rate_death"

#make wide data
usdrugdeaths<- usdrugdeaths %>%
  pivot_wider(names_from = cause_name, values_from = rate_death)

#change column names
colnames(usdrugdeaths)[colnames(usdrugdeaths)=="Alcohol use disorders"] <-"Alcohol"
colnames(usdrugdeaths)[colnames(usdrugdeaths)=="Opioid use disorders"] <-"Opioid"
colnames(usdrugdeaths)[colnames(usdrugdeaths)=="Cocaine use disorders"] <-"Cocaine"
colnames(usdrugdeaths)[colnames(usdrugdeaths)=="Amphetamine use disorders"] <-"Amphetamine"

#make line graph with all 4 drugs
drugdeaths <- ggplot(usdrugdeaths, aes(x = year)) +
  geom_line(aes(y = Alcohol, color = factor("Alcohol")), linetype = "solid", size = 1.65) +
  geom_line(aes(y = Opioid, color = factor("Opioid")), linetype = "solid", size = 1.65) +
  geom_line(aes(y = Cocaine, color = factor("Cocaine")), linetype = "solid", size = 1.65) +
  geom_line(aes(y = Amphetamine, color = factor("Amphetamine")), linetype = "solid", size = 1.65) +
  scale_color_manual(name = "Cause", values = c("Alcohol" = "dodgerblue", "Opioid" = "firebrick", "Cocaine"= "darkgreen", "Amphetamine"="purple")) +
  labs(title = "Drug-related Deaths in the US", 
       subtitle =  "Data from 1990-2019",
       caption = "Source: Institute for Health Metrics and Evaluation",
       x = "Year",
       y = "Death Rate per 100,000",
       color = "Cause") +
  scale_y_continuous(limits = c(0, max(usdrugdeaths$Alcohol, usdrugdeaths$Opioid, usdrugdeaths$Cocaine, usdrugdeaths$Amphetamine) + 1)) +
  theme_minimal() +
  theme(plot.title = element_text(color = "#0099f9", size = 20, face = "bold", hjust = 0.5),
        plot.subtitle = element_text(size = 13, face = "bold", hjust = 0.5),
        plot.caption = element_text(face = "italic", hjust = 0)) +
  theme(axis.title.x = element_text(color = "#0099f9", size = 16, face = "bold"),
        axis.title.y = element_text(color = "#0099f9", size = 16, face = "italic"))

print(drugdeaths)

# Convert ggplot to plotly
interactive_plot <- ggplotly(drugdeaths)

# View the interactive plot (can be exported as HTML)
interactive_plot